Gemini LLM Responses 10 Endpoints

Query Google Gemini through DataForSEO's AI Optimization API. Gemini is covered by two sub-APIs: LLM Responses (structured output) and LLM Scraper (raw scraped results from Google's AI-powered search).

Endpoints

Gemini LLM Responses (4 endpoints)

Call Name	Method	Endpoint
Set Gemini Tasks	POST	/v3/ai_optimization/gemini/tasks_post
Get Completed Tasks	GET	/v3/ai_optimization/gemini/tasks_ready
Get Re-parsed Tasks	GET	/v3/ai_optimization/gemini/tasks_fixed
Task GET	GET	/v3/ai_optimization/gemini/task_get/{id}

Gemini LLM Scraper (6 endpoints)

The Gemini scraper retrieves actual results from Google's AI-powered search interface (formerly Bard, now Gemini in Search).

Call Name	Method	Endpoint
Set Gemini Scraper Tasks	POST	/v3/ai_optimization/gemini/llm_scraper/task_post
Get Completed Scraper Tasks	GET	/v3/ai_optimization/gemini/llm_scraper/tasks_ready
Get Re-parsed Scraper Tasks	GET	/v3/ai_optimization/gemini/llm_scraper/tasks_fixed
Task GET Scraper	GET	/v3/ai_optimization/gemini/llm_scraper/task_get/{id}
Live Scraper	POST	/v3/ai_optimization/gemini/llm_scraper/live
Locations	GET	/v3/ai_optimization/gemini/llm_scraper/locations

Method

Gemini LLM Responses: Standard (async) — POST to set task, GET to retrieve.

Gemini LLM Scraper: Supports both Standard and Live. Use Live for instant results (single query). Use Standard for bulk/batch work.

Use Cases

• Determine whether a brand is cited in Gemini's AI-generated answers
• Compare Gemini recommendations to ChatGPT and Claude for the same query
• Track which competitor domains Gemini links to in its responses
• Monitor Gemini's shopping and local recommendations for a product category

LLM Responses Example

import os
import requests
from requests.auth import HTTPBasicAuth

auth = HTTPBasicAuth(os.environ["DATAFORSEO_LOGIN"], os.environ["DATAFORSEO_PASSWORD"])
BASE = "https://api.dataforseo.com"

# Post task
r = requests.post(
    f"{BASE}/v3/ai_optimization/gemini/tasks_post",
    auth=auth,
    headers={"Content-Type": "application/json"},
    json=[{
        "keyword": "who are the best local roofers in Dallas?",
        "location_code": 2840,
        "language_code": "en"
    }]
)
task_id = r.json()["tasks"][0]["id"]
print(f"Task: {task_id}")

LLM Scraper Example (Live)

r = requests.post(
    f"{BASE}/v3/ai_optimization/gemini/llm_scraper/live",
    auth=auth,
    headers={"Content-Type": "application/json"},
    json=[{
        "keyword": "roofing contractor Dallas TX",
        "location_code": 2840,
        "language_code": "en"
    }]
)
result = r.json()["tasks"][0]["result"][0]
print(result.get("response_text", "No response"))

Response Fields

Field	Type	Description
keyword	string	Query submitted
response_text	string	Gemini's full text response
brands_mentioned	array	Brands detected in response
urls_mentioned	array	URLs cited
se_type	string	"gemini"

Pricing

~$0.01 per LLM Responses task. LLM Scraper pricing varies — Live method costs more than Standard. Check account dashboard.

Notes

• Gemini is the only LLM provider with both a Responses API and a Scraper API — the scraper returns what users actually see in Google's AI-powered search UI
• The scraper is more representative of real AI Overview results than the structured Responses API
• Run llm_scraper/live for instant competitive checks; use task_post for bulk scheduled pulls